Open AccessCoal Price Forecast Based on ARIMA Model
Author(s) -
Xiaofan Zhang,
Chao Liu,
Yuhang Qian
Publication year - 2021
Publication title -
financial forum
Language(s) - English
Resource type - Journals
ISSN - 2251-2659
DOI - 10.18282/ff.v9i4.1530
Subject(s) - autoregressive integrated moving average , econometrics , coal , series (stratigraphy) , time series , computer science , economics , engineering , machine learning , paleontology , biology , waste management
This paper analyzes and determines the decision variables and constraints, establishes the EECM-ARAMA model to analyze and research coal price forecasts. Firstly, we first confirm the influencing factors. Then, we conduct correlation coefficient tests on price and various factors, and get the strength of the correlation between each factor and price. The second is to establish a coal price prediction model. Firstly, we use the EEMD method to transform the original price series into a stable time series, and then formulate three ARIMA models by comparing the size of the influencing factors and the parameter estimation results. After testing, we finally choose the ARIMA 03 model to predict the next 31 days, 35 Weekly and 36-month coal prices. Finally, we combine the models and ideas of the above issues to obtain factors that affect coal price changes and related price prediction models, and combine experience to give some feasible policy recommendations.